preprocess/humanparsing/run_parsing.py

import pdb
from pathlib import Path
import sys
import os
import onnxruntime as ort
PROJECT_ROOT = Path(__file__).absolute().parents[0].absolute()
sys.path.insert(0, str(PROJECT_ROOT))
from parsing_api import onnx_inference
import torch


class Parsing:
    def __init__(self, gpu_id: int):
        # self.gpu_id = gpu_id
        # torch.cuda.set_device(gpu_id)
        providers = [("CUDAExecutionProvider", {"device_id": torch.cuda.current_device(),
                    "user_compute_stream": str(torch.cuda.current_stream().cuda_stream)})]
        
        session_options = ort.SessionOptions()
        session_options.graph_optimization_level = ort.GraphOptimizationLevel.ORT_ENABLE_ALL
        session_options.execution_mode = ort.ExecutionMode.ORT_SEQUENTIAL
        # session_options.add_session_config_entry('gpu_id', str(gpu_id))
        # self.session = ort.InferenceSession(os.path.join(Path(__file__).absolute().parents[2].absolute(), 'checkpoints/humanparsing/parsing_atr.onnx'),
        #                                     sess_options=session_options, providers=['CPUExecutionProvider'])
        # self.lip_session = ort.InferenceSession(os.path.join(Path(__file__).absolute().parents[2].absolute(), 'checkpoints/humanparsing/parsing_lip.onnx'),
        #                                         sess_options=session_options, providers=['CPUExecutionProvider'])
        

        #new changes - 
        session_options.add_session_config_entry('gpu_id', str(gpu_id))
        self.session = ort.InferenceSession(os.path.join(Path(__file__).absolute().parents[2].absolute(), 'checkpoints/humanparsing/parsing_atr.onnx'),
                                            sess_options=session_options, providers=providers)
        self.lip_session = ort.InferenceSession(os.path.join(Path(__file__).absolute().parents[2].absolute(), 'checkpoints/humanparsing/parsing_lip.onnx'),
                                                sess_options=session_options, providers=providers)


    def __call__(self, input_image):
        # torch.cuda.set_device(self.gpu_id)
        parsed_image, face_mask = onnx_inference(self.session, self.lip_session, input_image)
        return parsed_image, face_mask